Calcium handling in skeletal muscles

Not only does Ca²⁺ play a pivotal role in the initiation, termination of muscle contraction, but it also serves as a key regulator in other cellular processes. Extensive research aiming at deciphering the complexity of the cascade regulating Ca²⁺ handling in skeletal muscle cells has been carried out during the last decades. This work led to the unveiling of several receptors, channels, transporters and carriers, which are responsible for the handling of this essential second messenger. These are discussed below.

1.1.3.1 Voltage-operated Ca²⁺ channels or dihydropyridine receptors

DHPR mediate the release of the intracellular Ca²⁺ from the SR following the depolarization wave initiated at the neuromuscular junction. Structurally, it is a complex of 5 subunits: alpha 1, alpha 2,

9 beta, gamma, delta (Catterall, 1991). DHPRs of fast twitch fibres were found to have 3 to 5-fold greater density than those of slow twitch fibres (Renganathan et al., 1998).

1.1.3.2 Ryanodine receptors

RyRs are large, high conductance Ca²⁺ channels that control the release of Ca²⁺ from the SR. The name was derived from the plant alkaloid ryanodine, due to the receptor’s ability to bind it with high affinity and specificity, which distinguishes them from another intracellular Ca²⁺ release channel family, the inositol 1,4,5-trisphosphate receptors (IP3R). Mammalian tissues express different RyRs isoforms: RyR1 is the dominant isoform in skeletal muscle, RyR2 is found in high levels in cardiac muscle, RyR3 is ubiquitously expressed, and all three isoforms are present in smooth muscles. RyRs are homotetramers constituted of 4 transmembrane helices forming a Ca²⁺

channel (Inui et al., 1987) and a very large protruding cytoplasmic domain, the foot region, which allows a direct connection between the channel and the DHPR at the plasma membrane (Brini &

Carafoli, 2000; Serysheva et al., 2005). Interaction between DHPRs and RyRs following an action potential occurs by protein-protein interaction (Paolini et al., 2004; Yin et al., 2005). RyRs are regulated by protein kinases and reactive oxygen species as indicated by the presence of phosphorylation sites (Hain et al., 1994), and a large number of free sulfhydryl groups, respectively (Hidalgo et al., 2005).

1.1.3.3 IP3R

IP3Rs are important regulators of Ca²⁺ release from the SR in smooth and cardiac muscle (Estrada et al., 2001); they are hardly expressed in skeletal muscle (Talon et al., 2002). However, they have been reported to be expressed in skeletal myoblasts and myotubes in culture (Jaimovich et al., 2000).

1.1.3.4 SR and sarcoplasmic reticulum Ca²⁺ ATPase

SR is the main store for Ca²⁺ in skeletal muscle with Ca²⁺ concentration as high as 1 mM at rest.

SERCA utilizes ATP mostly generated by the cytosolic isoform of creatine kinase (CK) bound to the SR to pump Ca²⁺ from the cytosol back into the SR against the Ca²⁺- concentration gradient (Korge et al., 1993; Rossi et al., 1990). SERCAs are encoded by three genes and alternative splicing leads to five different isoforms, the expression of which is tissue specific (Wuytack et al., 1992). SERCA1 is expressed only in skeletal muscles and shows a high abundance in fast twitch fibres, whereas SERCA2a is expressed in the heart and in slow twitch fibres. SERCA density is 5-7 times higher in fast

10 twitch compared to slow twitch fibres, to comply with the fast kinetics of Ca²⁺ removal needed by these muscles (Leberer & Pette, 1986).

1.1.3.5 Plasma membrane Ca²⁺ - ATPase (PMCA) and Na⁺- Ca²⁺ exchanger (NCX)

Pumping back Ca²⁺ into the SR is not the only pathway for extruding Ca²⁺ from the cytosol following contraction. Plasma membrane Ca²⁺-ATPase (PMCA) and Na⁺-Ca²⁺ exchanger (NCX) serve as extrusion pathways for Ca²⁺. While PMCA uses ATP to extrude Ca²⁺, NCX exchanges Ca²⁺ for Na⁺ (Strehler & Zacharias, 2001). However, their role is minor compared to SERCA in removing cytosolic Ca²⁺ after a contraction in healthy skeletal muscles.

1.1.3.6 Store operated Ca²⁺ channels (SOCs)

These are also referred to as capacitive Ca²⁺ entry pathways and are constituted by Ca²⁺ release activated Ca²⁺ channels (CRAC). These channels are found in most cells and are activated by SR Ca²⁺

depletion (Parekh & Putney, 2005). Abnormal function of such channels has been implicated in autoimmune diseases and inflammation. Activation of this Ca²⁺ entry pathway has been thoroughly investigated leading to the proposal of 2 mechanisms of activation. The first one is thought to occur through physical interaction between the channel-forming protein Orai1, and a sensor protein in the SR called Stim1. Upon store emptying, Stim1 oligomerises and interacts with the pore forming unit Orai1 leading to the activation of Ca²⁺ influx (Cahalan, 2009). The second one is thought to be due to release of a diffusible messenger termed calcium influx factor (CIF), which activates the Ca²⁺ -independent form of phospholipase A2 (iPLA2) by releasing the inhibitory protein calmodulin. This cascade is reported to activate SOCs through the production of lysophospholipids (Bolotina, 2008).

1.1.3.7 Stretch activated channels (SACs)

Stretch activated channels (SACs) are non-selective cationic channels that belong to the class of mechano-sensitive channels. Several members of the TRP channels have been described to be mechano-sensitive but to date the molecular identity of SAC in skeletal muscle have not been discovered (Holle & Engler, 2011). Activation of SACs is not fully understood; however, recent reports suggest the involvement of the microtubule network that conveys the mechanical signal subsequently to activate NADPH oxidases (NOXes) leading to the eventual opening of the channel (Khairallah et al., 2012; Prosser et al., 2012). These channels are inhibited by gadolinum ion (Gd³⁺), streptomycin, an aminoglycoside, and the tarantula spider (Grammostola spatulata) venom, grammotoxin (GsMTx-4) (Hamill & McBride, 1996).

11 1.1.3.8 Mitochondrial channels and transporters

Mitochondria play an essential role not only in energy production through oxidative phosphorylation, generation of reactive oxygen species (ROS) and programmed cell death, but are also involved in Ca²⁺ homeostasis due to their Ca²⁺ buffering capability (Pozzan & Rizzuto, 2000).

Ca²⁺ uptake by mitochondria is mediated by the uniporters, which is a gated channel driven by the electrochemical gradient across the inner mitochondrial membrane (Gunter et al., 2000). Other pathways of Ca²⁺ entry into mitochondria include the 'rapid-mode' uptake (RaM) (Sparagna et al., 1995) and the RyR1 localized to the inner membrane of mitochondria (Beutner et al., 2001). On the other hand, Ca²⁺ efflux is mediated by its exchange with Na⁺ or H⁺ through the antiporters (Carafoli, 2003). Another means of extruding not only Ca²⁺, but any molecule up to 1.5 kDa from the mitochondria is the mitochondrial permeability transition pore (mPTP). The mPTP structure has been shown recently to be formed from dimers of ATP synthase (Giorgio et al., 2013). Prolonged mPTP opening causes massive swelling of mitochondria, rupture of the outer membrane and release of inter-membrane components that induce apoptosis (Rasola & Bernardi, 2007).

1.1.3.9 Ca²⁺ buffering proteins

Several Ca²⁺-binding proteins exist in different cellular compartments such as calmodulin and parvalbumin in the cytosol, or calsequestrin and calreticulin in the SR. Most of these proteins belong to the EF-hand protein family and participate in Ca²⁺ handling by acting as Ca²⁺ buffering systems. One of the most important ones is calmodulin, which besides its ability to buffer Ca²⁺, interacts with a variety of other proteins and regulates their function (Vetter & Leclerc, 2003).

Another cytosolic protein, parvalbumin, is expressed in murine fast fibres but is absent in slow fibres and is completely absent in humans (Campbell et al., 2001). Calsequestrin is the major Ca²⁺

storage protein in the SR of all striated muscles (reviewed in Yano and Zarain-Herzberg (1994)). An illustration showing the pathways involved in Ca²⁺ handling in skeletal muscles is depicted in figure 1.5.

12

Figure 1-5, Calcium handling in skeletal muscle cells

The illustration shows different mediators of Ca²⁺ influx and extrusion pathways in skeletal muscle. The two compartments, which play a major role in Ca²⁺ buffering, the sarcoplasmic reticulum and the mitochondria, are shown in orange (adapted from Orrenius et al. (2003).

13 1.2 Duchenne muscular dystrophy (DMD)

“Muscular dystrophy” originates from the Latin word “dys”, meaning difficult and the Greek work

“trophe”, relating to nourishment, and denotes a heterogeneous group of neuromuscular disorders. Such disorders affect skeletal muscles and are accompanied by muscle wasting and weakness of variable distribution and severity, which may also affect cardiac and smooth muscles or other tissues (Engel et al., 1994). Inheritance in these disorders is either dominant or recessive, although the gene may be defective because of a de novo mutation (Arahata, 2000). Historically, the classification of these disorders were based on the phenotype till the early 1990’s when the identification of the genes involved changed the phenotype-based classification and shed new light on the molecular pathogenesis of these disorders (Cohn & Campbell, 2000). Currently, neuromuscular disorders are subdivided into several groups: Duchenne and Becker, Emery-Dreifuss, distal, facioscapulohumeral, oculopharyngeal and limb-girdle, which is the most heterogeneous group (Emery, 2002).

1.2.1 Clinical features and genetic aspects of DMD

DMD is an X-linked neuromuscular disease that affects about 1 in 3500 males and which results in progressive muscle degeneration (Emery, 2002). DMD patients typically present clinically at 4-5 years of age, with symptoms including difficulties in standing from a seated position, climbing stairs, and keeping up with their peers during play. The muscle degeneration is progressive and replacement with fibrous/fatty tissue occurs, which ultimately results in death by cardiac or respiratory failure by the late teens to the mid-twenties. With adequate supportive treatment, life expectancy has increased into late-twenties and early thirties (reviewed in (Bushby et al., 2010a, b).

Severe cognitive impairments such as reduced verbal skills are seen in approximately 30% of DMD cases (Moizard et al., 2000). Elevated levels of creatine kinase derived from muscle, in serum and amniotic fluid (Emery, 1977) remain the most widely used clinical markers in DMD diagnosis.

Diagnosis is confirmed by mutational analysis based on DNA sequencing tests.

The disease is caused by the lack of the protein dystrophin. The dystrophin gene is a large gene of 2.5 Mb transcribed to a 14 kb mRNA and coding for a 427 kDa protein (Hoffman et al., 1987). The large size of the gene, amounting to about 1% of the human X-chromosome, is a likely reason for the high rates of mutation (Hoffman & Schwartz, 1991). The gene is composed of 79 exons that map the Xp21 and occupies approximately 0.1% of the entire genome (Monaco & Kunkel, 1988).

About 70% of the affected boys inherit the gene from their carrier mother; the others exhibit

de-14 novo mutations, whereas their mothers have a normal X-chromosome. Approximately 60% of dystrophin mutations are large insertions or deletions that lead to downstream frameshift errors, whereas approximately 40% are point mutations or small frameshift rearrangements (Hoffman &

Dressman, 2001). This protein is localized just under the plasma membrane of muscle cells and links the extracellular matrix with the cytoskeleton. Several proteins are associated with dystrophin and are described in the following section.

1.2.2 Dystrophin and dystrophin-associated proteins complex

Dystrophin is a cytoskeletal linker protein linking F-actin to laminin-2 of the extracellular matrix. It is a rod-shaped protein consisting of four domains: an N-terminal actin-binding domain, twenty four triple helix spectrin-like repeats separated by four hinge regions, a cysteine-rich domain containing two predicted calcium binding motifs, and a unique C-terminal domain (Koenig et al., 1988).

Dystrophin localization is restricted to sub-sarcolemmal bands distributed along differentiated muscle fibres, where it binds F-actin at its N-terminus, and proteins such as nitric oxide synthase (NOS) via the syntrophins, the dystrobrevins, and syncoilin at its C-terminus (Suzuki et al., 1994).

The most important link regarding DMD is that made between the cysteine rich domain and the dystrophin glycoprotein complex (DGC) or dystrophin associated protein complex (DAPC).

Dystrophin appears to coordinate a link from the actin cytoskeleton to laminin-2 of the extracellular matrix (Ahn & Kunkel, 1993; Campbell, 1995) by way of β-dystroglycan (Jung et al., 1995), the sarcoglycans and sarcospan located in the sarcolemma, and α-dystroglycan on the external face of the membrane (Figure 1.6) (Ervasti & Campbell, 1991; Yoshida & Ozawa, 1990). Mutations in dystrophin or of any component of the DGC complex cause various types of muscular dystrophies (reviewed in Durbeej and Campbell (2002)). Yet the mechanism by which a protein absence causes a defined phenotype is still to be resolved awaiting a thorough knowledge of the biochemical and structural function of each member of the DGC.

15

Figure 1-6, Dystrophin associated protein complex

The illustration shows the central role orchestrated by dystrophin in linking not only the cytoskeletal F-actin with laminin in the extracellular matrix, but also to other essential proteins such as nitric oxide synthase and caveolin-3 (from Davies and Nowak (2006).

1.2.3 Pathogenesis of DMD

Despite the discovery of the primary defect in DMD, the secondary molecular mechanisms leading ultimately to muscle degeneration are diverse and interconnected. Evidence established from research on DMD patients and animal models points towards a crucial role for the loss of cytoskeletal and sarcolemmal integrity, deregulation of calcium homeostasis, enhanced protease activity, increased oxidative stress and impaired energy metabolism in the pathogenesis of the disease (Blake et al., 2002; Hopf et al., 2007). Historically, two hypotheses were presented in the literature to explain the pathogenesis seen in DMD: The structural/mechanical hypothesis and the dystrophin associated-signalosome hypothesis, however, there is no longer a clear distinction between the two.

In normal muscle, the DGC complex confers mechanical stability to the plasma membrane during muscle contraction by linking the cytoskeleton and the extracellular matrix, yet in DMD this stabilization effect is lacking and lead to tearing of the membrane during muscle exercise (Allen et al., 2005; Petrof et al., 1993). Evidence for the existence of such tears came not only from the presence of muscle-specific proteins in the blood of DMD patients, such as creatine kinase, but also

16 from the entry of membrane impermeable dyes, like Evans blue and Procion orange into dystrophic muscle cells when these were exposed to repeated contractions. It has also been reported that Ca²⁺

would enter extensively through these tears, leading to its chronic increment that activates a variety of downstream effectors ending up in muscle cell death.

It has also been reported that dystrophin and the DGC have several signalling functions and that loss of the complex in muscular dystrophy has been associated with alterations in multiple signal transduction pathways such as the mitogen-activated protein kinase (MAPK) or the Akt signalling pathways (Dogra et al., 2008). The biochemical repercussions of dystrophin absence in skeletal muscle cells are diverse, interconnected and amplify each other in viscous cycles. A description of the major downstream effectors of dystrophin absence is given below.

1.2.3.1 Ca²⁺ dysregulation

The lack of dystrophin has also been proposed to be directly responsible for enhanced Ca²⁺ entry in dystrophic muscle (Carlson, 1998; Hopf et al., 2007; Turner et al., 1991), because of abnormal activity of the non-selective cationic channels activated either by membrane stretch (SAC) or by Ca²⁺-store depletion (SOC) (Ducret et al., 2006; Vandebrouck et al., 2006) or through membrane tears (McNeil & Khakee, 1992). Recent evidence that support the central role played by Ca²⁺

dysregulation in the pathogenesis of DMD came from overexpressing TRPC3 channels, which was sufficient to replicate muscular dystrophy to some extent (Millay et al., 2009). Another study showed that the inhibition of Ca²⁺ influx by the expression of a dominant negative form of TRPV2 ameliorates muscular dystrophy in models of DMD (Iwata et al., 2009).

Several studies have shown that SOC activity is increased in dystrophic fibres (Boittin et al., 2006;

Harisseh et al., 2013). It was also reported that these channels are regulated by iPLA2 in dystrophic fibres and that the increased activity of this enzyme is likely responsible for the enhanced SOC activity in dystrophin-deficient fibres (Boittin et al., 2010).

SAC activity has also been reported to be increased in models of DMD. Recent studies show that activation of SAC channels is initiated by stretch that is conveyed through the microtubular network to activate NOXes, eventually activating these channels. This report established that the microtubule network is denser in mdx fibres, and that interventions antagonizing such an increment correct the increased Ca²⁺ influx through SACs (Khairallah et al., 2012; Prosser et al., 2012).

17 1.2.3.2 Protease activation

Calcium-dependent cysteine proteases, in particular -calpain, known to degrade a range of skeletal muscle proteins, including cytoskeletal and plasma membrane proteins, show both, elevated concentrations and activity in mdx muscles at the peak of muscle necrosis (Spencer et al., 1995). Calpain activity has been found elevated in extracts of muscle biopsies from DMD patients (Reddy et al., 1986). Despite the well-documented increase in calpain activity in DMD, and much efforts in research that have been invested, its role in the pathogenesis of muscular dystrophies is still unclear (Childers et al., 2011) and is probably of little importance.

1.2.3.3 Neuronal nitric oxide synthase (nNOS)

Loss of dystrophin in DMD results not only in delocalization of nNOS from the sarcolemma to the cytosol (Brenman et al., 1995), but also to a reduction of NO release from dystrophic muscle and decrement in circulating NO levels. Reduced NO levels lead to decreased superoxide scavenging and reduced vasorelaxation, local muscle ischemia and to an impairment of regeneration (Niebroj-Dobosz & Hausmanowa-Petrusewicz, 2005).

1.2.3.4 Ca²⁺ independent phospholipase A2 (iPLA2)

As stated above, iPLA2 plays a role in modulating Ca²⁺ entry not only through SOC but also through SAC. The first report on the role of iPLA2 in DMD came from Lindahl et al. on muscle biopsies in which a 10-fold increase of PLA2 activity as compared to controls was observed (Lindahl et al., 1995). Also, work from our lab showed increased expression of iPLA2 in mdx muscles (Boittin et al., 2006).

This enzyme is also known as group VIA PLA2, and was initially purified and characterized from the P388D1 macrophage-like cell line (Ackermann et al., 1994). Group VIA PLA2 has a molecular weight of around 85 kDa, and can exist in an aggregated form. It contains multiple ankyrin repeats, which may play a role in its oligomerisation. It exhibits lysophospholipase activity as well as phospholipase A2 activity, and is capable of hydrolysing a wide variety of phospholipid substrates. iPLA2 catalyses the hydrolysis of phospholipids at the sn-2 position, releasing a free fatty and a lysophospholipid.

Two isoforms are reported to be present in skeletal muscles, the short and the long isoform (Winstead et al., 2000). The long isoform has a 54 amino acid insert, which disrupts the 8^th ankyrin repeat (figure 1.7) and is reported to mediate membrane association (Hsu et al., 2009).

18

Figure 1-7, Structure of group VIA-2 phospholipase A2

A, Schematic representation of the different domains of iPLA2. B, Deuterium exchange map showing the 7 ankyrin repeats and the catalytic domain (Hsu et al. 2009).

1.2.3.5 Oxidative stress

Oxidative stress is generally defined as an imbalance that favours the production of ROS over cellular antioxidant defences. Early studies have suggested that increased oxidative stress contribute to muscle damage and degeneration in dystrophic muscle. Protein carbonyls and lipid peroxidation by-products, markers of oxidative stress, have been detected in muscles of DMD patients and mdx mice (Murphy & Kehrer, 1986; Ragusa et al., 1997; Rodriguez & Tranopolsky, 2003).

The demonstration that oxidative stress is not only a secondary effect of muscle degeneration but precedes the pathologic changes of muscle dystrophy, came from the observation that lipid peroxidation products and expression of antioxidant enzymes were increased in 3-week-old mdx mice before the onset of the disease (Disatnik et al., 1998).

Sources of oxidative stress in respiratory and locomotor muscles in DMD are thought to include mitochondria, NOXes, and infiltrating inflammatory cells, and on the other hand, reduced ROS

19 scavenging due to delocalization of nNOS (reviewed in Lawler (2011)). In the current overview, we will focus on mitochondria and NOXes due to their pertinent role in the oxidative stress seen in DMD.

1.2.3.5.1 Mitochondria

Mitochondria have generally been cited as the predominant source of ROS in muscle cells. Early reports suggested that 2-5% of the total oxygen consumed by mitochondria might undergo a one-electron reduction with the generation of superoxide, and that complexes I and III of the one-electron transport chain are the main sites of mitochondrial superoxide production (Davies et al., 1982;

Koren et al., 1983). During exercise, a number of researchers have assumed that the increased ROS generation that occurs during contractile activity is directly related to the elevated oxygen consumption that occurs with increased mitochondrial activity. Such a correlation would imply a 50- or 100-fold increase in superoxide generation by skeletal muscle during aerobic contractions (Kanter, 1994). However, recent findings suggest that NOXes and not mitochondria are the dominant source of ROS during exercise (Sakellariou et al., 2013b; Xia et al., 2003).

1.2.3.5.2 NADPH oxidases

The NOX family members are transmembrane proteins that transport electrons across biological

The NOX family members are transmembrane proteins that transport electrons across biological